← Search

Rafał Szlendak

2 accepted papers

2024

Understanding Progressive Training Through the Framework of Randomized Coordinate Descent

AISTATS 2024poster

We propose a Randomized Progressive Training algorithm (RPT)—a stochastic proxy for the well-known Progressive Training method (PT) (Karras et al., 2017). Originally designed to train GANs (Goodfellow et al., 2014), PT was proposed as a heuristic, with no convergence analysis even for the simplest o…

Cited by 1Source ↗PDF ↗Scholar ↗
2022

Permutation Compressors for Provably Faster Distributed Nonconvex Optimization

ICLR 2022poster

In this work we study the MARINA method of Gorbunov et al (ICML, 2021) -- the current state-of-the-art distributed non-convex optimization method in terms of theoretical communication complexity. Theoretical superiority of this method can be largely attributed to two sources: a carefully engineered…

Cited by 44Source ↗PDF ↗Scholar ↗